The present invention relates to wireless communication. The present invention is especially applicable to mobile wireless communication and processing of signals to compensate for nonlinearities associated with power amplifiers used for wireless transmission.
Next-generation wireless communication will utilize improved transmitter technology for variety of broadband and multimedia services, supported by advanced potable equipment and handsets. For a long time, the longer handset battery life-time, better call quality and less wireless high-frequency radiation to human brain are three most concerning and insolvable issues for mobile phone designers and users. Actually, the above three problems are closely related to spectral efficiency and power efficiency of handset, because both higher spectral efficiency and power efficiency will greatly improve mobile system performance, extend handset battery life-time and reduce handset transmitted power to some extent. However, the nonlinear distortion and low DC conversion efficiency introduced by power amplifier (PA) in current handset RF transmitter impact severely performance of wireless system and shortens greatly battery life-time of handset.
Next-generation wireless communication will require improved transmitter technology for variety of broadband and multimedia applications, supported by advanced improved base stations and access points to potable equipment and handsets. The spectral efficiency and power efficiency are among the most important requirements of mobile communication systems. For many years, the designers of wireless communication system have been concerning the issue of PA linearization in RF transmitter, because it is closely related to the development of highly spectral efficiency modulation scheme. It has been demonstrated that the spectrally efficient linear modulation technologies such as QPSK and QAM have high spectral efficiency under the case of linear amplification. However, in order to obtain the highest power efficiency, the nonlinear power amplifier such as Class AB, C or D is required in final amplifying stage of RF transmitter. Unfortunately, the high power efficiency of nonlinear amplifiers generates nonlinear inter-modulation products in adjacent channels, which results in both amplitude to amplitude (AM-AM) and amplitude to phase (AM-PM) distortion. The nonlinear distortion will cause spectral broadening and high out-of-band power emission of output signal. As the result, the signal spectrum expands into adjacent channels to produce interference for other users. Furthermore, these inter-modulation distortion products are spaced so close to the desired signal and cannot simply be filtered out by conventional filters. In order to avoid the nonlinear distortion, a simple and usual solution is to back off output signal from the saturation range of PA so that signal level is confined to the power amplifier. However, this will result in a less power efficient operation because several decibels of back off are required usually to obtain appropriate linearity. Obviously, the approach is not suitable to the advanced wireless system design that should be a high capacity and efficient digital transmission system.
For a long time, the longer handset battery's life-time, less radio radiation to human brain and high voice quality have been the most concerning problems by handset designers, makers and users. Actually, the key to solve above problems is closely related to develop the handset RF system with spectral efficiency and power efficiency. The higher spectral efficiency and power efficiency not only improve greatly mobile system performance, but also extend battery life-time of handset and reduce transmitter's radiation which will be helpful to protect users' brain from radiation to some extent.
Current predistortion technologies used widely to linearize PA in mobile communication system are mainly analog predistorter implemented at IF/RF by means of analog circuit and digital predistorter at baseband with digital signal processing (DSP) technique.
The analog predistorter is based on the principle of error subtraction and power match to realize linearization of PA, and, hence must use an auxiliary PA to match the main PA. Under a perfect matching, the error of auxiliary PA will compensate nonlinear distortion caused by main PA. Because nonlinear feature of PA is very complicated and many variables are involved, the analog predistortion has only less predistortion accuracy and consumes more power.
In contrast, the DSP-based predistorter is usually preferred since it has stable characteristics that perform in a wide range of temperatures, and eliminates the necessary of tuning in factory. Therefore, it is better suitable to the fast tracking and adjusting any possible changes in PA parameters, such as drifts due to temperature, aging and operating point variations.
Unfortunately, although DSP-based predistortion technologies have advantages over analog ones, it is difficult for current digital predistortion schemes to apply to handset and wideband system because of complexity in hardware organization and DSP algorithm.